Search for Amazon PPC tips and you'll read the same advice a hundred different ways. Keep your ACoS low. Add budget to what's working. Let the algorithm learn. Check in once a week.
Follow every word of it and you can still end up with a shrinking business. We've opened plenty of accounts where the metrics look respectable, the campaigns look tidy and the brand is losing ground to competitors who understood something the tips never mention.
Most of that advice isn't wrong, exactly. It just stops short of the part that costs real money.
Checking ACoS and nudging bids is not management
Here's what "managing" PPC looks like in most accounts we inherit. Log in. Glance at ACoS. Nudge a few bids down. Pause anything that looks scary. Log out until next week.
That's not optimising. It's babysitting.
Real optimisation asks harder questions. Which search terms are converting and which are just collecting clicks. Whether the exact-match campaign is cannibalising the phrase campaign sitting next to it. Which products still make money after fees and ad spend, and which only look like they do.
Those questions take an afternoon, not a glance. They're also where the money is. A bid nudge changes what you pay for a click. The harder questions change whether the click was worth buying at all.
TACoS includes the part that pays your bills
TACoS, total advertising cost of sale, is ad spend divided by total revenue. Paid and organic together.
Watch it over months and it tells you something ACoS never can. If your TACoS is falling while spend holds steady, your organic business is growing underneath the ads, which is exactly what you want. The advertising is building an asset. If your TACoS is flat or rising, you're renting your sales. Turn the ads off and the revenue goes with them.
Two brands, both at 25% ACoS. One has a TACoS of 8% and falling. The other sits at 24%, because almost nothing sells without a sponsored placement. Same ad efficiency on paper. Completely different businesses.
The low-ACoS trap
A low ACoS feels like winning. Often it means you're under-investing.
If your ACoS is sitting at 8% and your break-even is 25%, that gap is growth you decided not to buy. Nobody hands out prizes for the spend you didn't make. Every impression you didn't pay for went to a competitor, and on Amazon rank compounds: their sales feed their organic position, which feeds more sales.
We'd go further. Chasing the lowest possible ACoS is one of the fastest ways to shrink an Amazon business. You retreat to branded terms and cheap long-tail, the dashboard looks brilliant, and your reach to new customers falls off a cliff. Six months later organic rank has slid and nobody in the business can work out why, because every weekly report said the ads were "performing".
ACoS on its own can't tell you any of this, which is why we look at total revenue against total ad spend instead. We've written up the full argument in our piece on ACoS vs TACoS.
Budget rarely fixes what a match type would
When sales stall, the reflex is to raise the budget. But budget only amplifies what's already in the account. If a broad-match campaign is dragging in traffic that doesn't convert, more budget buys more of that traffic, faster.
Before adding a pound, look at structure. Most stalled accounts don't need more traffic. They need the traffic they're already buying sorted into campaigns that can actually be read — proven terms in one place, exploration in another, broad match on a much shorter leash. In most accounts, that restructure does more than an extra £50 a day ever would.
And check who else is spending on your keywords. If unauthorised resellers are sitting on your listings, some of them are running ads on your own brand terms, which means you're in an auction against people selling your own stock. No bid strategy fixes that. That's a channel-control problem wearing a PPC costume, and we've covered why it matters in our piece on why your Amazon price keeps falling.
Measure contribution margin, not the dashboard
ACoS and ROAS describe the relationship between ad spend and ad revenue. Neither tells you whether you made any money.
Contribution margin does. Take the sale price, subtract product cost, Amazon's referral fee, fulfilment, storage and the ad spend it took to win the order. What's left is the number the business actually lives on. Plenty of "hero" products with a lovely ACoS turn out to be roughly break-even once the full stack of fees comes off. Plenty of unloved products turn out to be quietly funding everything else.
This is why we run ring-fenced ad budgets tied to margin rather than to mood. One brand we manage runs £10,000 a month in ad spend, agreed up front, sized against what the products genuinely earn per unit. The budget doesn't jump because someone got nervous or excited. It moves when the margin maths says it should.
Where Amazon PPC tips stop and management starts
When a brand comes to us complaining that ads "aren't working", the account usually tells the same story. Auto campaigns untouched for two years. No negative keywords anywhere. Branded and generic spend mixed into one campaign so nobody knows what's driving what. No idea which products make money after fees.
None of that gets fixed by a tip. It gets fixed by someone treating the account like it's their own money on the line.
That's the real problem with most Amazon PPC tips: they treat advertising as a dial you turn, separate from everything else. In practice your ads sit on top of your price, your content, your reviews and who else is selling on your listing. If those are broken, the ads inherit the breakage. Most brands assume they've got a marketing problem when what they've actually got is a control problem, and no bid strategy in the world spends its way past that.
If your account feels busy but the business feels stuck, the ads are usually trying to tell you something the ACoS column can't. That's the sort of thing a second pair of eyes finds inside a week, and we'll tell you what we see even if the answer is that you don't need us.
.png)
.png)
.svg){kind=icon}
